[go: up one dir, main page]
More Web Proxy on the site http://driver.im/

CN106331678A - Hardware-implemented floating-point type viewpoint processing method in glasses-free 3D display system - Google Patents

Hardware-implemented floating-point type viewpoint processing method in glasses-free 3D display system Download PDF

Info

Publication number
CN106331678A
CN106331678A CN201610833722.4A CN201610833722A CN106331678A CN 106331678 A CN106331678 A CN 106331678A CN 201610833722 A CN201610833722 A CN 201610833722A CN 106331678 A CN106331678 A CN 106331678A
Authority
CN
China
Prior art keywords
viewpoint
floating
sub
pixel
estimation
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Granted
Application number
CN201610833722.4A
Other languages
Chinese (zh)
Other versions
CN106331678B (en
Inventor
秦剑飞
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Sichuan Changhong Electric Co Ltd
Original Assignee
Sichuan Changhong Electric Co Ltd
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Sichuan Changhong Electric Co Ltd filed Critical Sichuan Changhong Electric Co Ltd
Priority to CN201610833722.4A priority Critical patent/CN106331678B/en
Publication of CN106331678A publication Critical patent/CN106331678A/en
Priority to PCT/CN2017/095502 priority patent/WO2018054173A1/en
Application granted granted Critical
Publication of CN106331678B publication Critical patent/CN106331678B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Classifications

    • HELECTRICITY
    • H04ELECTRIC COMMUNICATION TECHNIQUE
    • H04NPICTORIAL COMMUNICATION, e.g. TELEVISION
    • H04N13/00Stereoscopic video systems; Multi-view video systems; Details thereof
    • H04N13/10Processing, recording or transmission of stereoscopic or multi-view image signals
    • H04N13/106Processing image signals
    • H04N13/122Improving the 3D impression of stereoscopic images by modifying image signal contents, e.g. by filtering or adding monoscopic depth cues

Landscapes

  • Engineering & Computer Science (AREA)
  • Multimedia (AREA)
  • Signal Processing (AREA)
  • Testing, Inspecting, Measuring Of Stereoscopic Televisions And Televisions (AREA)
  • Image Processing (AREA)

Abstract

The invention relates to a glasses-free 3D display technology, discloses a hardware-implemented floating-point type viewpoint processing method in a glasses-free 3D display system, and aims to perform floating point processing and error compensation with high efficiency and high precision under the condition of hardware implementation so as to satisfy the display effects of different glasses-free 3D screens when the grating tile angle and tile distance change. According to the invention, the viewpoint value of a first sub pixel in each row is calculated precisely; then, the integer parts and floating point parts of the viewpoint values of other sub pixels are respectively arranged and estimated, the estimation result is compensated, and a cycle even distribution law is adopted in estimation and compensation; and finally, the final viewpoint values are calculated based on the integer parts and compensated floating point parts of the viewpoint values of the sub pixels.

Description

Can hard-wired floating type viewpoint processing method in bore hole 3D display system
Technical field
The present invention relates to bore hole 3D Display Technique, be specifically related in bore hole 3D display system to regard by hard-wired floating type Point processing method.
Background technology
In recent years, ordinary consumer improves year by year for bore hole 3D TV demand, and therefore bore hole 3D television cost is constantly While ground reduces, also begin to the bore hole 3D television system that various implementation different editions occurs.From full SOC system to entirely GPU system, the most is that full GPU system realizes.In order to reduce cost, application A8 chip adds the side that simple FPGA controls Method, but the picture display effect obtained is less desirable, and cost minimizing is also far smaller than expected value.
Bore hole processing scheme on existing market can be divided into integer and floating type, integer scheme to lose major part signal, draws Face display effect is very poor;The complexity realized due to algorithm of floating type and amount of calculation are too big, the most only it Middle very small part realizes for hardware, does not also have a whole set of hard-wired method.
If a whole set of bore hole 3D system whole flow processs after rgb signal enters all are realized with hardware, bore hole 3D electricity To be substantially reduced depending on the cost of machine, average family is bought bore hole 3D TV and will be possibly realized.
Summary of the invention
The technical problem to be solved is: for the defect of prior art, proposes a kind of bore hole 3D display system In can hard-wired floating type viewpoint processing method, it is intended to efficient and high-precision float in the case of hardware is attainable Point processes and error compensation, with the satisfied different bore hole 3D screens display effect when grating slope angle and slant range change.
The technical solution adopted for the present invention to solve the technical problems is:
In bore hole 3D display system can hard-wired floating type viewpoint processing method, comprise the following steps:
A, parameter according to bore hole 3D screen, the viewpoint value of first sub-pixel of a line every to picture calculates;
B, viewpoint value N to first sub-pixel of every a line are cast out process and are obtained its integer part;
C, integer part to the N of the sub-pixel position of every a line are estimated;
D, floating-point part to the N of the sub-pixel position of every a line are averaged arranged evenly and carry out floating-point weight and estimate Calculate;
E, the error bringing viewpoint estimation carry out the backward compensation with the regularity of distribution;
F, according to estimation viewpoint integer part and compensate after floating-point weight portion calculate final viewpoint value.
As optimizing further, in step A, to the method for the viewpoint value calculating of first sub-pixel of every a line it is:
According to seeking viewpoint formula: N=Mout/P*VviewCalculate viewpoint value, the viewpoint value asked for is retained eight floating-points; Wherein, N is viewpoint value, MoutMove up and down the computational constant of relation for display screen physical resolution and grating slope, P is grating Physical parameter, VviewNumber for viewpoint.
As optimizing further, in step C, the integer part of the N of the sub-pixel position of every a line is arranged in order and by Estimate according to the rule moved in circles.
As optimizing further, in step D, the method for described floating-point weight estimation is:
Starting estimation from second sub-pixel of every a line, floating-point is increased by sub-pixel position, and increments is to zero, then From zero increase to this journey, the value of first sub-pixel, so circulates, until the floating-point weight of last sub-pixel position of this journey Estimation completes.
As optimizing further, in step E, the described error bringing viewpoint estimation carries out the backward benefit with the regularity of distribution Repay, specifically include:
The floating-point part cast out is needed to be defined as F in addition to eight floating numbers retained the viewpoint value asked in step A, will The sub-pixel of co-located, according to the mode of distribution same with the mode that the is evenly distributed backward in step D, is floated by the F of first sub-pixel Point weight compensates, and obtaining final floating-point part is NF
Floating-point weight portion as optimizing further, in step F, after the described viewpoint integer part according to estimation and compensation Point calculate final viewpoint value method be:
By formula NEventually=NZ*NF+(NZ+1)*(1-NF) calculate the viewpoint value that final upper screen multiple views merges.
The invention has the beneficial effects as follows: improve the efficiency of algorithm, floating type bore hole 3D can be realized with low cost hardware and show Show system, cost-effective.
Detailed description of the invention
Bore hole 3D algorithm in the present invention, from the beginning of the image data of HDMI or LVDS path inputs, each Step all hardware realizes, and utilizes DDR3 storage data to realize interframe and switches, the restructuring of picture and image quality post-processing algorithm, the most firmly Part realizes.What difficulty was maximum among these is exactly the integer process of floating type bore hole 3D screen and intercepts the process of certain precision floating point data The hardware of algorithm realize, and the error compensation hardware after processing for corresponding floating-point realizes the optimization of algorithm, specifically walks Rapid as follows:
1, to sub-pixel (x, y) calculating of accurate viewpoint value:
Here x represents input picture lateral coordinates position, and y represents input picture longitudinal coordinate position.According to eye 3D screen Parameter (grating slope angle, grating paster slant range), to input the every a line of picture each RGB data carry out viewpoint Calculate, the formula according to seeking viewpoint: N(x,y)=Mout/P*VviewCalculating viewpoint value, it must be a decimal, its fractional part Divide i.e. viewpoint weight to take precision 1/256, i.e. retain eight floating numbers;
This calculating comprises the complex calculation such as multiplication and division, and it is many that Hardware Implementation takies resource, but the inventive method only calculates Every first sub-pixel of a line, other overwhelming majority sub-pixel need not calculate, and direct result makes to realize by hardware exactly Bore hole 3D algorithm become to achieve with low cost hardware;In above formula, N is viewpoint value, MoutFor display screen physical resolution with Grating slope moves up and down the computational constant of relation, and P is grating physical parameter, VviewNumber for viewpoint.
2, first sub-pixel calculated viewpoint value N of every a line is cast out process and obtain its integer part, i.e. N Deduct integer value N that the floating-point part of N obtainsZ
3, the integer part of the N of the sub-pixel position of every a line is estimated, such as: if this picture is to use 8 to regard Point merges, then the mode according to { 1,2,3,4,5,6,7,8,1,2,3,4,5,6,7,8 ... } arranges estimation;
4, arranging the floating-point part of the N of the sub-pixel position of every a line, rule is by by first sub-pixel of this journey The floating-point part being calculated N is averaged arranged evenly;
5, the calculating to the floating-point viewpoint weight of the N of the sub-pixel position of every a line: open from second sub-pixel of every a line Begin estimation, floating-point increase by sub-pixel position and increments to zero, then the value of first sub-pixel from zero increase to this journey, as This circulation, until last sub-pixel position of this journey has been estimated;
6, the viewpoint value of estimation is compensated: by the N value of calculating in step 1 in addition to eight floating numbers retained, will be super Cross 8, need the floating-point partial value cast out to be defined as F, when specifically compensating, by the F of first sub-pixel according to in step D It is evenly distributed the mode backward mode with distribution, the sub-pixel floating-point weight of co-located is compensated, obtains final floating-point Part is NF
7, by formula NEventually=NZ*NF+(NZ+1)*(1-NF) calculate the viewpoint value that final upper screen multiple views merges.
Comparison software of the present invention calculates the method for viewpoint, obtains result and has error slightly, but combines regarding of human eye Feel feature, error control in the receptible scope of human eye, employing can hard-wired system architecture, improve the effect of algorithm Rate, allows the bore hole 3D displaying scheme cost of costliness be reduced significantly.

Claims (6)

1. can hard-wired floating type viewpoint processing method in bore hole 3D display system, it is characterised in that comprise the following steps:
A, parameter according to bore hole 3D screen, the viewpoint value of first sub-pixel of a line every to picture calculates;
B, viewpoint value N to first sub-pixel of every a line are cast out process and are obtained its integer part;
C, integer part to the N of the sub-pixel position of every a line are estimated;
D, floating-point part to the N of the sub-pixel position of every a line are averaged arranged evenly and carry out floating-point weight estimation;
E, the error bringing viewpoint estimation carry out the backward compensation with the regularity of distribution;
F, according to estimation viewpoint integer part and compensate after floating-point weight portion calculate final viewpoint value.
2. in bore hole 3D display system as claimed in claim 1 can hard-wired floating type viewpoint processing method, its feature It is, in step A, to the method for the viewpoint value calculating of first sub-pixel of every a line is:
According to seeking viewpoint formula: N=Mout/P*VviewCalculate viewpoint value, the viewpoint value asked for is retained eight floating-points;Wherein, N For viewpoint value, MoutMove up and down the computational constant of relation for display screen physical resolution and grating slope, P is grating physics ginseng Number, VviewNumber for viewpoint.
3. in bore hole 3D display system as claimed in claim 2 can hard-wired floating type viewpoint processing method, its feature It is, in step C, the integer part of the N of the sub-pixel position of every a line is arranged in order and enters according to the rule moved in circles Row estimation.
4. in bore hole 3D display system as claimed in claim 3 can hard-wired floating type viewpoint processing method, its feature Being, in step D, the method for described floating-point weight estimation is:
Starting estimation from second sub-pixel of every a line, floating-point is increased by sub-pixel position, and increments is to zero, then from zero It is incremented to the value of first sub-pixel of this journey, so circulates, until the floating-point weight estimation of last sub-pixel position of this journey Complete.
5. in bore hole 3D display system as claimed in claim 4 can hard-wired floating type viewpoint processing method, its feature Being, in step E, the described error bringing viewpoint estimation carries out the backward compensation with the regularity of distribution, specifically includes:
The floating-point part cast out is needed to be defined as F, by first in addition to eight floating numbers retained the viewpoint value asked in step A The sub-pixel floating-point of co-located, according to the mode of distribution same with the mode that the is evenly distributed backward in step D, is weighed by the F of individual sub-pixel Heavily compensating, obtaining final floating-point part is NF
6. in bore hole 3D display system as claimed in claim 5 can hard-wired floating type viewpoint processing method, its feature Being, in step F, the floating-point weight portion after the described viewpoint integer part according to estimation and compensation calculates final viewpoint value Method is:
By formula NEventually=NZ*NF+(NZ+1)*(1-NF) calculate the viewpoint value that final upper screen multiple views merges.
CN201610833722.4A 2016-09-20 2016-09-20 It can hard-wired floating type viewpoint processing method in naked eye 3D display system Active CN106331678B (en)

Priority Applications (2)

Application Number Priority Date Filing Date Title
CN201610833722.4A CN106331678B (en) 2016-09-20 2016-09-20 It can hard-wired floating type viewpoint processing method in naked eye 3D display system
PCT/CN2017/095502 WO2018054173A1 (en) 2016-09-20 2017-08-01 Hardware-implemented floating-point viewpoint processing method in glasses-free 3d display system

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN201610833722.4A CN106331678B (en) 2016-09-20 2016-09-20 It can hard-wired floating type viewpoint processing method in naked eye 3D display system

Publications (2)

Publication Number Publication Date
CN106331678A true CN106331678A (en) 2017-01-11
CN106331678B CN106331678B (en) 2019-04-30

Family

ID=57787462

Family Applications (1)

Application Number Title Priority Date Filing Date
CN201610833722.4A Active CN106331678B (en) 2016-09-20 2016-09-20 It can hard-wired floating type viewpoint processing method in naked eye 3D display system

Country Status (2)

Country Link
CN (1) CN106331678B (en)
WO (1) WO2018054173A1 (en)

Cited By (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN107483912A (en) * 2017-08-31 2017-12-15 成都斯斐德科技有限公司 A kind of multi-view image fusion method based on floating type lenticular sheet
WO2018054173A1 (en) * 2016-09-20 2018-03-29 四川长虹电器股份有限公司 Hardware-implemented floating-point viewpoint processing method in glasses-free 3d display system
CN111104091A (en) * 2019-12-12 2020-05-05 北京科技大学 Detection and conversion method for precision specific calculation in dynamic floating point error analysis

Citations (7)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN102708577A (en) * 2012-06-15 2012-10-03 中航华东光电有限公司 Method for synthesizing multi-view-point three-dimensional picture
CN103813153A (en) * 2014-01-27 2014-05-21 北京乐成光视科技发展有限公司 Weighted sum based naked eye three-dimensional (3D) multi-view image synthesis method
CN104796685A (en) * 2015-03-31 2015-07-22 王子强 Universal free three-dimensional display system
CN104811686A (en) * 2015-04-14 2015-07-29 西安交通大学 Hardware implementation method for floating-point type multi-view naked-eye three-dimensional synthetic image
US9110298B2 (en) * 2012-04-04 2015-08-18 JVC Kenwood Corporation Naked-eye stereoscopic display apparatus
CN105338336A (en) * 2015-10-28 2016-02-17 四川长虹电器股份有限公司 Lenticular grating and LCD screen-based stereo image fast synthesis method
CN105898284A (en) * 2015-11-24 2016-08-24 徐州维林苑文化传媒有限公司 Continuous viewpoint mirror-free three-dimensional grating parameter superiority judging method

Family Cites Families (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN106331678B (en) * 2016-09-20 2019-04-30 四川长虹电器股份有限公司 It can hard-wired floating type viewpoint processing method in naked eye 3D display system

Patent Citations (7)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US9110298B2 (en) * 2012-04-04 2015-08-18 JVC Kenwood Corporation Naked-eye stereoscopic display apparatus
CN102708577A (en) * 2012-06-15 2012-10-03 中航华东光电有限公司 Method for synthesizing multi-view-point three-dimensional picture
CN103813153A (en) * 2014-01-27 2014-05-21 北京乐成光视科技发展有限公司 Weighted sum based naked eye three-dimensional (3D) multi-view image synthesis method
CN104796685A (en) * 2015-03-31 2015-07-22 王子强 Universal free three-dimensional display system
CN104811686A (en) * 2015-04-14 2015-07-29 西安交通大学 Hardware implementation method for floating-point type multi-view naked-eye three-dimensional synthetic image
CN105338336A (en) * 2015-10-28 2016-02-17 四川长虹电器股份有限公司 Lenticular grating and LCD screen-based stereo image fast synthesis method
CN105898284A (en) * 2015-11-24 2016-08-24 徐州维林苑文化传媒有限公司 Continuous viewpoint mirror-free three-dimensional grating parameter superiority judging method

Cited By (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
WO2018054173A1 (en) * 2016-09-20 2018-03-29 四川长虹电器股份有限公司 Hardware-implemented floating-point viewpoint processing method in glasses-free 3d display system
CN107483912A (en) * 2017-08-31 2017-12-15 成都斯斐德科技有限公司 A kind of multi-view image fusion method based on floating type lenticular sheet
CN111104091A (en) * 2019-12-12 2020-05-05 北京科技大学 Detection and conversion method for precision specific calculation in dynamic floating point error analysis
CN111104091B (en) * 2019-12-12 2021-11-26 北京科技大学 Detection and conversion method for precision specific calculation in dynamic floating point error analysis

Also Published As

Publication number Publication date
WO2018054173A1 (en) 2018-03-29
CN106331678B (en) 2019-04-30

Similar Documents

Publication Publication Date Title
CN102572486B (en) Acquisition system and method for stereoscopic video
CN106448608B (en) Screen brightness adjusting method and device aiming at mura problem and television
CN105244007B (en) A kind of generation method and device of the grayscale correction chart of camber display screen
CN103260081B (en) A kind of video image zooming processing method and processing device
CN106331678A (en) Hardware-implemented floating-point type viewpoint processing method in glasses-free 3D display system
US20110134315A1 (en) Bi-Directional, Local and Global Motion Estimation Based Frame Rate Conversion
CN102231099A (en) Method for correcting per-pixel response brightness in multi-projector auto-stereoscopic display
CN102332158B (en) Image synthesis method and system using alpha channels
CN102254348A (en) Block matching parallax estimation-based middle view synthesizing method
US9578312B2 (en) Method of integrating binocular stereo video scenes with maintaining time consistency
CN104811686A (en) Hardware implementation method for floating-point type multi-view naked-eye three-dimensional synthetic image
US20180270400A1 (en) Liquid crystal display device and image processing method for same
WO2011109898A1 (en) Generating 3d multi-view interweaved image(s) from stereoscopic pairs
CN104296876A (en) FPGA-based scene non-uniformity correction method and device
CN102547338A (en) DIBR (Depth Image Based Rendering) system suitable for 3D (Three-Dimensional) television
CN104182950B (en) Image processing method and device thereof
CN105338336B (en) Lenticular grating and LCD screen-based stereo image fast synthesis method
CN105405389A (en) Method and system for correcting bright color degree of LED display device, and method for correcting regional bright color degree
CN109284062B (en) Touch data processing method, device, terminal and medium
CN103024247A (en) Electronic image stabilization method based on improved block matching
CN109612391A (en) Based on the binocular measuring device of separate unit high speed camera to land leveller method for space coordinates measurement
CN102186095B (en) Matching error correction method applicable for depth-image-based rendering
CN103002230B (en) Adaptability pixel compensation method
CN103686029B (en) A kind of video pixel opposite-angle interpolation method for the many optical projection systems of giant-screen
CN102724528A (en) Depth map generation device

Legal Events

Date Code Title Description
PB01 Publication
PB01 Publication
C10 Entry into substantive examination
SE01 Entry into force of request for substantive examination
CB03 Change of inventor or designer information

Inventor after: Qin Jianfei

Inventor after: Jiang Zhiqiang

Inventor after: Han Daqiang

Inventor before: Qin Jianfei

CB03 Change of inventor or designer information
GR01 Patent grant
GR01 Patent grant